This is an RO1 proposal to study the genetics of cocaine dependence using a collection of small nuclear families, each including (at least) an affected sibling pair (ASP). The clinical work will take place at four university-based programs in CT (Univ. CT and Yale), MA (McLean Hospital), and SC (Med. Univ. of South Carolina), and the laboratory and statistical work at Yale and Boston University School of Medicine. This project will provide information that will, eventually, substantially increase understanding of the mechanisms of cocaine dependence, and lead to new ways to attack this pervasive societal problem therapeutically. The purpose of this proposal is to identify chromosomal regions containing genes predisposing to cocaine dependence (CD). The aims are to collect a set of 500 small nuclear families, primarily ASPs, and (whenever possible) additional siblings and parents; and to complete a 10 cM genome scan using STR markers and an automated sequencer. Affection will be defined according to DSM-IV diagnostic criteria, ascertained using the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA) as the primary instrument. The principal method of analysis will be sibling pair linkage; this will be augmented by linkage disequilibrium methods. An important genetic contribution to risk for CD is strongly supported by clinical genetic data. This sample will have sufficient power to detect linkage under reasonable assumptions of heterogeneity. As for any complex trait, while the proposed sample size cannot be proven a priori to be sufficient to complete gene mapping, it is necessary, and the cell lines prepared as part of this study may be a component of any future mapping efforts. These data, combined with mapping data from other substance dependence (SD) disorders, will ultimately help to resolve basic questions about the extent of unique, vs. shared, genetic contribution to various SD phenotypes. This will be the first large-scale linkage study of CD. Given that CD is a complex trait, a large ASP study that will allow genetic analysis without assumptions about trait transmission will provide the best chance to map genes related to this dangerous, prevalent disorder.